Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/CodeAbra/iai-personal-memory-engineWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/codeabra/iai-personal-memory-engine/iai-directive)<a href="https://agentmods.dev/commands/codeabra/iai-personal-memory-engine/iai-directive"><img src="https://agentmods.dev/badge/commands/codeabra/iai-personal-memory-engine/iai-directive/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/codeabra/iai-personal-memory-engine/iai-directive"><img src="https://agentmods.dev/badge/commands/codeabra/iai-personal-memory-engine/iai-directive.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00017 | $0.00334 |
| Opus 5.5 | $0.00007 | $0.00134 |
| Sonnet 5.5 | $0.00003 | $0.00067 |
| Haiku 4.5 | $0.00002 | $0.00033 |
Grade B, and why
iai-directive scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Unrestricted tool accessmediumExcessive agency
A wildcard tool grant or "run any command" leaves no least-privilege boundary at all.
If `$ARGUMENTS` is `remove <id>`, do not run any command and do not send What it actually says
This command has three forms, dispatched on $ARGUMENTS. Do not
paraphrase or summarize the user's text before passing it; forward it
verbatim as the quoted argument. Each form is a thin wrapper over its own
CLI/chat mechanism -- it adds no privilege that mechanism does not already
grant.
Set a standing order
If $ARGUMENTS is not list and does not start with remove , run the
following command exactly, substituting the user's typed rule for
$ARGUMENTS, and report its output:
iai capture --directive "$ARGUMENTS"
List live standing orders
If $ARGUMENTS is exactly list, run:
iai directive list
and report its output verbatim.
Remove a standing order by id
If $ARGUMENTS is remove <id>, do not run any command and do not send
the removal marker yourself -- only a genuine turn the user themselves
types can retire a directive. Instead, tell the user to type (or paste)
the following line as their own next message, substituting the given
<id>:
remove directive: <id>
This mirrors how setting a directive relies on the user's own typed
standing directive: prefix -- the removal marker must likewise
originate from the user's own message, never from an assistant reply.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago Changed · +30 lines · -1 tokens per session scan A → B e27f160f65f5
- 1mo ago First seen · 18 lines · 18 tokens per session scan A 422aaf8c687e
iai-directive is a command published in the GitHub repository CodeAbra/iai-personal-memory-engine (913 stars, last pushed yesterday), licensed MIT. It adds 17 tokens to every session and 334 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it B with 1 finding (unrestricted tool access). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-05.
Other commands, from other repositories
weekly
Weekly memory report — facts learned, procedures, repeated mistakes prevented.
reason
Think through a question using memory-backed retrieval.
recall
Semantic + metadata search across stored drawers.
remember
Store a verbatim memory drawer in the current wing.
log
Command "log" from allenc84/sapience, covering /log — judgment ledger command, routing, new assessment: /log, review: /log review [domain] and resolve: /log resolve.
recall
Answer from reef memory, reading the pages before answering.